Alcohol use disorders are extremely common in bipolar disorder and, when present, are associated with increased rates of hospitalization, poor outcomes, violence towards self and others, and non- adherence to bipolar disorder treatment. However, minimal data are available on the treatment of these patients. The current application is in response to PA-13-160 (Alcohol Use Disorders: Treatment, Services, and Recovery Research) that emphasizes medications development, the need to Discover, develop, and test new, more effective agents to prevent or reduce drinking, and the Treatment for Psychiatric/Substance Abuse/Medical Comorbidity, as well as the need to Develop and implement new, efficient adaptive clinical trial designs and statistical analyses. We propose an innovative design that will examine two promising novel pharmacotherapies (citicoline and pregnenolone) in bipolar disorder and alcohol use disorders. Citicoline has potent neuroprotective properties in animal models and appears to decrease both cocaine and alcohol use in humans. Pregnenolone is a precursor to GABAergic neurosteroids. In addition to effects on GABA and other neurotransmitters, pregnenolone enhances microtubule assembly, a pathway essential for neuroplasticity that is impaired in chronic alcohol use. Pregnenolone reduces alcohol consumption in animal models. Thus, both compounds have novel properties that are pertinent to alcohol dependence but that have not yet been explored in clinical trials. In addition to the preclinical and clinical literature supporting these agents for mood symptoms and alcohol dependence, we present pilot data supporting their use in patients with bipolar disorder and alcohol dependence. An innovative adaptive trial design called drop-the-loser will be implemented such that the trial will begin with two active treatment arms and a placebo arm. At interim analysis, a treatment arm can be dropped if ineffective. At this point, the study will continue as a two-arm trial comparing the more promising agent against placebo. If both agents seem effective then the trial will continue with three treatment arms. If neither agent appears promising at interim analysis then the trial will be discontinued. The approach allows for rapid assessment of multiple treatments as well as an adequately powered trial of one treatment. The study will be conducted at two sites using two principal investigators, who are highly experienced researchers in the field. An expert on adaptive clinical trials will conduct the statistical analysis. Two sites will maximize the use of resources needed to achieve specific aims in an efficient manner. The goal is both to develop a promising, and badly needed, treatment for bipolar disorder and alcohol use, and demonstrate the utility of an innovative clinical trial methodology in the alcohol treatment field.